|
Tårns Audio PA100 Power Stage
The Tårns Audio PA100 uses a relatively unconventional voltage amplification structure without a classical long tailed differential input pair. Instead, the amplifier operates around a single ended voltage amplification node where the input signal and the feedback signal are summed directly.
The input signal enters through the 2.2 µF coupling capacitor. The 47 kΩ resistor establishes the input impedance while the 2.2 kΩ resistor participates in the DC biasing of the first transistor stage. The 220 pF capacitor at the input acts as an RF suppression filter to improve ultrasonic stability.
The two transistors near the input form a complementary gain structure rather than a differential amplifier. This gives the amplifier a more direct signal path with fewer active devices in series with the audio signal.
The central node where the two 22 Ω resistors meet is extremely important because it acts as the main summing point of the amplifier.
The 39 kΩ resistor returns signal from the output back to this node. The 820 Ω resistor and 100 µF capacitor establish the lower leg of the feedback network.
This is therefore the global negative feedback loop.
The 100 µF capacitor makes the lower portion of the network frequency dependent. At audio frequencies the capacitor effectively bypasses much of the 820 Ω resistor, increasing AC feedback while preserving the desired DC operating conditions.
The approximate closed loop voltage gain can be estimated from:
A= approx 1 + 39K/820
which gives approximately:
Av= 48.5 that expressed in dB is 33.7
This is actually a fairly typical gain value for a power amplifier of this type.
Because the amplifier lacks a differential pair, the feedback is injected directly into the emitter referenced signal summing node. This is an older but elegant method that can work extremely well if transistor matching and thermal behavior are carefully controlled.
The two 22 Ω resistors provide local degeneration and help linearize the small signal stage. They also help stabilize current flow between the complementary halves of the voltage amplifier section.
The voltage amplification stage itself is formed by the central transistors associated with the 560 Ω, 1.5 kΩ, 180 Ω, and 4.7 Ω resistors. This section develops most of the amplifier’s voltage gain.
The compensation strategy is intentionally light. The 1 nF capacitors provide dominant pole stabilization while the 50 pF capacitor introduces high frequency lead compensation. The designer clearly intended the amplifier to operate with relatively wide bandwidth and high slew rate rather than extremely aggressive frequency rolloff.
This suggests the amplifier was designed to minimize transient intermodulation distortion and preserve fast signal transitions.
The dual 235 Ω potentiometer arrangement forms the bias spreader for the output stage. This establishes the VBE voltage across the driver chain and therefore controls idle current and crossover distortion.
The output stage is a classic EF3 topology:
voltage amplifier → predriver → driver → output transistor.
This triple emitter follower arrangement provides very large current gain while keeping the loading on the voltage amplification stage extremely low.
That is one of the reasons amplifiers like this can achieve very fast slew rates and good subjective transparency.
The output transistors are paralleled using 0.22 Ω emitter resistors to equalize current sharing and improve thermal stability.
With ±38 V rails, expected power output is likely around:
70 to 90 watts into 8 Ω
and approximately:
120 to 140 watts into 4 Ω
assuming adequate heatsinking and transistor SOA.
The output network consisting of the 1.8 µH inductor, the 10 Ω resistor, and the 100 nF Zobel capacitor is a classic high frequency stabilization system. These components isolate the amplifier from highly capacitive speaker loads and help prevent ultrasonic oscillation.
Because this amplifier is relatively fast and lightly compensated, PCB layout is critical. Poor grounding or excessive wiring inductance could easily produce ringing or parasitic oscillation.
Sonically, this type of amplifier will probably sound very immediate and dynamic. The relatively moderate feedback level combined with the high speed EF3 output structure usually produces:
excellent transient response, strong bass articulation, open midrange presentation, and a highly detailed top end.
Rather than chasing ultra low static THD numbers, the design philosophy appears focused on preserving dynamic linearity and minimizing transient artifacts.
Technically, it is a very elegant old school high speed bipolar power amplifier using moderate global feedback, a minimalist front end, and a fast triple emitter follower output stage.
|
